I am bad at electronics, help with ULN2803 darlington arrays... / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

I am bad at electronics, help with ULN2803 darlington arrays...

1 2 next

Move Thread

LAN_403

AndrewM

Mon Sept 29 2014, 05:07PM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

I can't figure out what is going on with this project!!!

I have the following circuit mocked up, IC1 and IC2 are ULN2803A's:

PspNspN

Basically I'm trying to switch my twelve loads (R_load) using 8x signals from my Arduino. To do so, I switch on either B1 or B2 to "select" the proper ULN2803, and then switch on the proper C1 through C6 line to select the load. I simulated this is Pspice and LTSpice and it works beautifully.

However nothing works right on the breadboard.

Initially I didn't have the 12x diodes present, because I didn't think I needed them. But in practice turning on any "C" line made every load on that ULN2803 turn on. So I added the diodes and that appeared to help...

... except now when I look at the voltage at point A, I see some influence from other channels:
W6GrhIS

None of this was expected in Pspice. Pulse width as shown is about 5us. I tried adding a pull down resistor to each base of the darlington, it lowers the height of the extraneous peaks, but doesn't eliminate them.

What am I doing wrong??

Steve Conner

Mon Sept 29 2014, 05:36PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I think what you're doing wrong is switching the ground pin of the ULN. :P

When the MOSFET is off and any of the control lines is high it will act like an emitter follower and pull its ground pin up to 5V less a few diode drops. This voltage will then feed back to all the other input pins through the ULN's internal pulldown resistors. It might look ugly, but it should work. I can't see any reason why it wouldn't. Inductive loads might be sketchy, I recommend connecting the ULN's diode pin to the positive rail to catch any spikes.

AndrewM

Mon Sept 29 2014, 05:55PM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

OK. Well.

a) the problem I'm having only occurs when the MOSFET is on.

b) if switching the ground is sketchy, how would you gang multiple ULN2803's together?

Steve Conner

Mon Sept 29 2014, 07:55PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

What is the problem then? You never specified. Does it fail to switch the loads somehow?

I would gang multiple ULN2803s by driving each one off an octal latch like the 74HC573, or an I2C to GPIO expander chip like the PCF8574. Or simply a bigger micro with more GPIO pins.

We use the PCF8574+ULN2803 combo at work for driving relays off I2C. Be aware that the PCF8574 can't source much current and needs pullup resistors.

If you look at the schematic on page 2 of the datasheet

you can see why the inputs would crosstalk when the MOSFET was off, but I can't see why they would crosstalk when the MOSFET was on. What MOSFET are you using? Is the Rds(on) low enough?

Patrick

Mon Sept 29 2014, 08:13PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

I dont see how this can work at all, maybe you Pspice'd it wrong?

As steve said, turning the mosfet off should cuase things to get pulled up to V+

AndrewM

Mon Sept 29 2014, 10:59PM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

Something is causes my loads to burn out (they're very sensitive to duty cycle), and I suspect the "extra" pulses I see are the culprit.

But.

I see what you're saying about the crosstalk with the mosfet off. I think I need to investigate further.

----------

Steve you nailed it... apparently it was only floating when the mosfet was off...

So now I have no idea why I blew my load.

Snicker.

Patrick

Tue Sept 30 2014, 03:24AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

what king of load are they?
if there in the linear mode, half-on, half-off maybe they die more easily.

mbd

Tue Sept 30 2014, 12:06PM

Registered Member #43278 Joined: Sat Feb 22 2014, 09:18AM
Location: Sydney
Posts: 9

Turning back to the OP, why did you decide to 'switch' the darlington arrays by controlling their ground pins? Wouldn't it have been more straightforward to use one array rather than two, in a multiplexed configuration, with two mosfets (or BJTs) switching the +24V side of the loads? (You would eliminate one array and the diodes on the inputs but would probably need diodes on the outputs to prevent back circuits.)

MBD

Steve Conner

Tue Sept 30 2014, 12:19PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The problem with that is that if you put the switches in the 24V rail, you need a high-side driver for them. Not really that complicated (PNP BJT switches would make it particularly simple) but not trivial either.

AndrewM

Tue Sept 30 2014, 04:22PM

Registered Member #49 Joined: Thu Feb 09 2006, 04:05AM
Location: Bigass Pile of Penguins
Posts: 362

More importantly, my loads have a common +24V rail. Thus my "solution."

I'm replacing all the blow stuff tonight and will give it another try to see if it blows again. As far as I can tell everything is working properly..........

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob